Optimal Web Service Selection Using UML Profile

Abstract

Enterprises are more conscious of providing quality of services over the web for reasons of economy, reliability, interoperability and flexibility. Enterprise application relies on selection of the most appropriate service from several candidate services with similar capabilities provided by different service providers. The question is, on what basis the system chooses a service among several candidates. This paper proposes a model that makes an automatic selection of best service and detects the variance between the non-functional requirements of the users and service qualifications. In this paper, we describe our approach aimed to detect conflicts between user requirements and the service specifications of the service provider. Our work proposes to detect these conflicts using Ontology and UML profiles to achieve better performance and avoid unpredictable state of the system. We suggest use of UML extensions and domain Ontology to detect NFR conflicts between the client’s requirements and service specifications.